Cerebras Q2 Revenue Up 74.3% to $180.11 Million, Shares Plunge 18%

Cerebras reported its financial results for the second quarter, with revenue up 74.3% to $180.11 million, but its shares plunged over 18% in after-hours trading as it missed analysts' expectations. The company's cloud services revenue surged 281%, but hardware sales dropped, leading to an operating expense increase and a loss of roughly $450.53 million.

Despite this, Cerebras raised its annual revenue and gross margin forecasts, citing strong demand for its AI chips. The company expects 2026 adjusted revenue to be between $880 million and $890 million, with an adjusted gross margin of 40.6%, down from 46.5% in the previous quarter.

In other news, Nebius reported a 514% surge in AI cloud sales, driving its shares up 20%. The company's cloud sales reached $514 million, with a total contract value of $4 billion. Snowflake's Observe platform now allows AI agents direct access to operational telemetry data, enabled by the updated Model Context Protocol (MCP) server and CLI tool.

Additionally, CrowdStrike expanded its Project QuiltWorks security framework to small and medium-sized businesses, combining AI-driven vulnerability discovery, remediation services, and cyber insurance protection. The AI Trust and Security Consortium (AITSC) was launched to set peer-defined standards for enterprise AI, bringing together cross-industry security and IT leaders to build AI guardrails.

Key Takeaways

• Cerebras Q2 revenue up 74.3% to $180.11 million, but shares plunged 18% due to missed expectations. • Cerebras expects 2026 adjusted revenue to be between $880 million and $890 million. • Nebius AI cloud sales surged 514% to $514 million. • Snowflake's Observe platform allows AI agents to access operational telemetry data. • CrowdStrike expanded Project QuiltWorks security framework to small and medium-sized businesses. • AI Trust and Security Consortium (AITSC) launched to set standards for enterprise AI. • Karoline Leavitt used AI-generated content in her farewell statement. • Elon Musk predicts AI revenue will outpace all other SpaceX business lines within months. • MobileJudgeBench benchmark introduced for evaluating LLM-based judges on mobile agent trajectories.

Cerebras shares plummet 20% after earnings miss

Cerebras reported its financial results for the second quarter, with revenue up 74.3% to $180.11 million, but its shares plunged over 18% in after-hours trading as it missed analysts' expectations. The company's cloud services revenue surged 281%, but hardware sales dropped. Cerebras' operating expenses rose, and it lost roughly $450.53 million.

Cerebras raises annual targets on strong AI chip demand

Cerebras Systems raised its annual revenue and gross margin forecasts, citing strong demand for its AI chips. The company's adjusted gross margin was 40.6%, down from 46.5% in the previous quarter. Cerebras expects 2026 adjusted revenue to be between $880 million and $890 million.

AI Trust and Security Consortium launches

The AI Trust and Security Consortium (AITSC) was launched to set peer-defined standards for enterprise AI. The consortium aims to bring together cross-industry security and IT leaders to build AI guardrails.

Karoline Leavitt used AI in exit statement

Karoline Leavitt, former White House Press Secretary, used AI-generated content in her farewell statement. The statement included an image that appeared to be AI-generated.

Nebius shares leap 20% on AI cloud sales

Nebius reported a 514% surge in AI cloud sales, driving its shares up 20%. The company's cloud sales reached $514 million, with a total contract value of $4 billion.

Observe connects AI agents to telemetry

Snowflake's Observe platform now allows AI agents direct access to operational telemetry data. The updated Model Context Protocol (MCP) server and CLI tool enable AI systems to query Observe's context graph.

CrowdStrike expands AI security program

CrowdStrike expanded its Project QuiltWorks security framework to small and medium-sized businesses. The program combines AI-driven vulnerability discovery, remediation services, and cyber insurance protection.

NCW Tech Alliance Artificial Intelligence Expo

Protestors gathered outside the Wenatchee Convention Center to protest AI, data centers, and companies sponsoring the NCW Tech Alliance's AI Expo. The protest was organized by activists concerned about the environmental impact and ethics of AI.

Today in AI: Musk's AI revenue bombshell

Elon Musk predicts AI revenue will outpace all other SpaceX business lines within months. AI infrastructure stocks surged, with CoreWeave and Super Micro gaining 19%.

Benchmarking LLM judges for mobile agent evaluation

Researchers introduced MobileJudgeBench, a benchmark for evaluating LLM-based judges on mobile agent trajectories. The study found that a simple baseline judge with sampled screenshots was competitive with purpose-built methods.

Sources

NOTE:

This news brief was generated using AI technology (including, but not limited to, Google Gemini API, Llama, Grok, and Mistral) from aggregated news articles, with minimal to no human editing/review. It is provided for informational purposes only and may contain inaccuracies or biases. This is not financial, investment, or professional advice. If you have any questions or concerns, please verify all information with the linked original articles in the Sources section below.

Cerebras AI chip demand revenue gross margin forecast expectation loss operating expenses AI Trust and Security Consortium standards enterprise AI security IT guardrails Karoline Leavitt AI-generated content farewell statement image Nebius AI cloud sales shares Observe AI agents telemetry Snowflake Model Context Protocol MCP server CLI CrowdStrike AI security program Project QuiltWorks vulnerability discovery remediation services cyber insurance protection NCW Tech Alliance Artificial Intelligence Expo protest environmental impact ethics Elon Musk AI revenue SpaceX CoreWeave Super Micro LLM MobileJudgeBench benchmark evaluation mobile agent trajectories judge screenshots purpose-built methods

Comments

Loading...